Periodontal disease (gingivitis and periodontitis), prediabetes and type 2 diabetes mellitus (T2DM) are important public health problems as they are highly prevalent and their treatment is both difficult and expensive. Type 2 Diabetes (T2D) and periodontal disease are known to be strongly correlated. A large body of evidence suggests these correlations are causal and bidirectional. While diabetes is a generally well- accepted risk factor for the development of periodontal disease, classical bidirectional hypotheses only suggest periodontal disease to contribute to poor glycemic control among patients with established diabetes. However, in this proposal we hypothesize that dysbiotic subgingival microbial communities common in periodontal disease (i.e., periodontal infections) might contribute to insulin resistance and future development of prediabetes among initially diabetes-free individuals. Recent work by our group and others provides strong support for this hypothesis. Nevertheless, several limitations exist limiting causal inference and obscuring future research directions. First, temporality is unclear due to cross-sectional designs. Second, no study to date has comprehensively assessed the subgingival microbiome as it relates to prediabetes. Third, the potential role of inflammation as a mediator of associations between subgingival microbiota and prediabetes has not been examined in longitudinal studies. We will enroll and follow longitudinally 1,400 diabetes-free individuals to examine the co-evolution of the subgingival microbiome and the development of prediabetes. This will allow us to test hypotheses about the potentially bidirectional association between periodontal infection and impaired glucose regulation early in the natural history of both disease processes and well before the development of overt diabetes. This investigation will also generate important knowledge necessary for optimizing the design of future intervention studies. Specifically, future RCT designs will be optimized by knowledge about microbial community characteristics and inflammatory phenotypes associated with prediabetes risk, to ensure that susceptible individuals are enrolled and anti-infective treatment goals are achieved.